First, they met the boy. He was 6 when he joined the school, struggling with problems at home, suffocated by the challenges of poverty. He was far behind his peers in all areas of learning and needed urgent help.

So, interventions were put in place, extra assistance was arranged, pastoral support was provided and outreach work with his family became extensive.

But there was too much lost time to make up and too little school time to do it. He got to Year 6 still behind where he should have been.

His sister was luckier: because they met the boy, they met her, too. She was only two years old. The school and the local authority worked with the parents, and worked with her, to improve the home environment, to provide structured interventions within the school, to spot the key issues and try to fix them. She came to school not just equal to her more privileged peers but having surpassed them.

This school (anonymous to protect the identity of the children) is not alone in reaching out before the usual starting age to the families of potential pupils.

It is increasingly accepted that, by Reception, the gap between disadvantaged children and their peers can be too great to close through schooling alone.

So, schools are taking over pre-schools, running community groups, offering outreach interventions, accepting ever-younger children into the school environment in whichever way they can. One school has even installed a soft-play area open to local families with under threes.

Policymakers around the world have seized on the idea, too: getting involved in the homes of struggling families has, for example, been a key part of the UK government’s social mobility plans.

The idea is simple and the path well-trodden: prevention is better than cure. And there is research to back up its application in education. Perhaps the most eminent advocate of early intervention programmes is James Heckman, the Nobel prize-winning University of Chicago economist, who has written extensively on the role of early intervention in boosting attainment. His theories propose that the earlier the intervention educationally, the higher the return later.

“High-quality early childhood education programmes enrich the learning and nurturing environments of disadvantaged children,” he writes in a 2015 paper. “An accumulating body of evidence shows the beneficial effects of these programmes.”

Tom McBride, director of evidence at the Early Intervention Foundation and a former head of strategic analysis at the Department for Education, agrees.

“You would expect that providing disadvantaged children with high-quality, well-organised, structured early education could make a difference and there is reasonable evidence, from a number of different studies, that it can,” he says.

And yet, the story is not as one-sided as it seems. “We certainly welcome that more attention and resources are dedicated to providing high-quality services in the early years for children at risk of atypical development, but how you guarantee quality and deliver it is a really complicated question,” says McBride. “And Heckman himself would say that, while such early intervention is necessary, it is not sufficient.

“It is no good providing really high-quality early-years support for disadvantaged children and then letting them go to terrible, dysfunctional schools with low aspirations, poor-quality teaching or poor leadership, and thinking that is going to reduce the disparity in outcomes in GCSEs or Sats.”

There is also widespread misunderstanding about the issues that are becoming the focus of interventions in this age group.

Language development is the hot topic of the moment but experts say it is unlikely to be helped by screening two-year-olds and putting a short-term intervention in place for those who present with issues.

We still don’t know, either, what type of intervention may be effective. A “fix” of any description may be impossible for some. And a whole range of language problems associated with disadvantage may not actually manifest until much later ages.

So where does this leave early intervention? What exactly does the research suggest that educators should - or can - do?

The sooner the better?

The government clearly thinks that early intervention is a good thing. Early-years provision and early intervention is an area that has been prioritised by the DfE as part of its aim to cut the attainment gap between disadvantaged youngsters and their peers.

One key strand of this is the expansion of free childcare for three- and four-year-olds, announced last year; and the DfE has also pledged, by 2028, to halve the number of children who start school without the speaking or reading skills they need. In announcing that aim in a speech at the Resolution Foundation in July, education secretary Damian Hinds was clear that concentrating interventions on the early years of life is crucial: “The point of greatest leverage for social mobility is the very earliest time in life - what happens between the ages of 0 and 3,” he said.

“Primarily, that means what happens at home…when you’re behind from the start you rarely catch up because, of course, your peers don’t wait, the gap just widens and this has a huge impact on social mobility.

“On average, disadvantaged children are four months behind at age 5. That grows by an additional six months by the age of 11 and a further nine months by the age of 16. So, by the time they take their GCSEs they are, on average, 19 months behind their peers.”

Efforts to close that gap include schools coming forward with “projects to create new high-quality nursery places” and training early-years workers so they have “the expertise to support very young children’s early language and literacy skills”.

McBride says intervening at this age makes sense. “If you are talking about reducing the disparity in cognitive development, or speech and language, we know these gaps open up before the age of 3, and you see them very clearly at the point children enter school. So there is a very clear theoretical basis to say you shouldn’t wait for these gaps to emerge and widen rather than try to tackle them as early as possible.”

Indeed, most researchers looking at childhood development are in agreement with Heckman that trying to help children who struggle with learning at an early age is, in theory, a good thing. However, many point out that knowing you should intervene and knowing how to do so are two very different things.

This is perhaps best demonstrated by looking closely at the development of language: it forms the basis of the government’s aims for intervention in early childhood and is a sobering demonstration of how complicated those aims might be to achieve.

For starters, how clear cut is the link between poverty and challenges in language, and is singling out poverty as a reason to get involved in a home - and using poverty as the basis for the formulation of any intervention - problematic? After all, we have seen many proposals about simply increasing the number of books in a home, or increasing vocabulary exposure, the assumption being that poverty equates to a lack of those things.

“For most child development issues, whether you are looking at social and emotional learning, or speech and language, or behaviour, then there will be a social gradient to these things,” argues McBride.

Theoretically, he says, there are two ways in which poverty would be impacting on children’s speech and language development. First is the direct impact of being poor, which can produce a more stressful environment and can impair carers’ ability to provide a supportive nurturing environment.

“It doesn’t guarantee they can’t, but if you are worried about how you are going to pay for the electricity or put food on the table, on average, this can make it more difficult,” says McBride.

Then there are indirect causes; the characteristics for which poverty can be a proxy measure.

“We have done work about behaviour in the past, finding that the association between economic disadvantage and childhood behavioural and emotional difficulties is explained by parental levels of education and maternal psychological wellbeing - so poverty acts as a proxy for a set of other factors that will impact upon children’s emotional development,” says McBride.

But Courtenay Norbury, professor of developmental language and communication disorders at University College London, argues it can be a much more complicated link than many believe.

“Heckman is probably right - high-quality and sustained nursery provision for disadvantaged children is the best bet for improving social mobility. But I would be cautious in suggesting that this would completely resolve all language problems before school entry.

“There are plenty of children from poor households who develop fantastic language skills. Nevertheless, there is a well-known association between socio-economic disadvantage and language disorder, and the reasons for this are complex,” she says.

Dorothy Bishop, professor of developmental neuropsychology at the University of Oxford, agrees. “Poverty doesn’t necessarily cause language problems in children - consider someone like Alan Johnson MP, who grew up in pretty desperate poverty and is highly articulate. There are plenty of examples like that. There’s a correlation, but causation is not demonstrated.”

She stresses that the influence of genetics cannot be ignored. “Given that language problems are familial, and adults with poor language skills tend not to get well-paid jobs, there’s a cycle of disadvantage and it’s difficult to disentangle genetic and environmental influences, which no doubt interact.

“We know that language development and disorder are shaped by genetic influences. We don’t fully understand this, but we think there are some genetic risk factors that affect the way a child’s brain develops in ways that are non-optimal for language learning.”

As detailed in our cover feature of 7 September, “Destiny’s New Assistant”, written by genetics researcher Kathryn Asbury, methods are being - and have been - developed to identify risk factors that may mean we can spot children who might be at risk of developing problems. But as that article expressed, doing so would be highly controversial.

Peter Scott, professor of child health and behaviour at King’s College London and director of the National Academy for Parenting Research, says that this genetic link to development is frequently overlooked - partly because it is still regarded as a taboo subject. “It is quite interesting that people will accept that for musical ability or running that genes matter but it is still considered slightly politically incorrect to say that some kids are better readers than others - or that they are more intelligent than others,” he says. “The narrative, and the journalism, haven’t caught up with the science on this.”

Complex issue

Unpicking all this complexity to find a cause of the development issues and an intervention that is effective is tricky. Norbury illustrates the issue with an example: a child with developmental language disorder (DLD).

“We know that DLD is a lifelong condition - although language skills will develop and improve over time, it is very likely that the child will become an adult who has language difficulties,” she says.

“And because language is so central to learning and reading, this adult with DLD is very likely to have left school with limited literacy and very few educational qualifications. This limits opportunities for employment and further education.

“Longitudinal studies show that while many adults with DLD can and do work and live independently, they tend to have lower-skilled jobs that don’t pay well.”

This problem, she continues, can become “a vicious cycle” - embedding itself in future generations. “Imagine that adult with DLD has a child of his or her own. That child likely has the same genetic vulnerabilities to develop DLD, but is now also in a low-income household.

“This child’s parents may not find talking or reading very easy, and may be less inclined to do so if the child isn’t responsive or initiating talk. Poverty comes with all sorts of social stressors that might take up parents’ time. They may well recognise that their child has the same sorts of language issues they had, but they don’t have the resources to seek extra help or fight for different education provisions.”

It is important to stress that this is not the case for all children and families affected by DLD, Norbury says - particularly since DLD can occur in very wealthy families, too. However, the research shows that it is more common where there is socio-economic disadvantage, and so it is important to recognise that, in these cases, what looks like an “environmental” effect can also reflect genetic influences on language and their interactions with the environment.

DLD is, of course, just one example of the complexity around challenges with language development. Other developmental problems have complexities of their own.

With that in mind, do we know anything that works for language challenges in younger age groups? “The first thing to say is that it would be great to have more robust research evidence investigating different types of early intervention - there is a lot we don’t know,” Norbury admits. “The second thing is that age is really important.”

One large trial, carried out by Melissa Wake and colleagues, from the Murdoch Children’s Research Institute in Melbourne, looked at the effects of providing a public health intervention for late talkers - children who were slow to say their first word or phrase.

For the trial, children were selected because their expressive language at age 2 was in the bottom 20 per cent, and they were randomly allocated to receive a parent language training programme or to a no-treatment control group. “The parents received six sessions of input from speech-language therapists about parent talk and ways to facilitate language development - which is pretty similar to what many speech-language therapy services will offer to under 5s in the UK,” Norbury explains.

The children were then followed up when they were aged 3.

“Parents loved the intervention, but there was no difference between the intervention and control groups on standard tests of language,” Norbury adds. “This was because the control group improved just as much as those receiving the intervention.”

Another study, led by the University of Melbourne’s Patricia Eadie, investigated short-term training programmes for nursery staff that aimed to improve children’s language. It did demonstrate improvements in how the staff communicated with the children, “but unfortunately, this did not result in improvements in children’s language”, Norbury says.

“There are probably many reasons why this is so: [it may be that] children in the control groups were improving as well; the short period of the study might not have been long enough to observe any impact on the children; and the nursery staff were not qualified teachers, so it may be that they require a longer training period to fully develop the knowledge and skills they need around developing child language. We [also] know that many two-year-olds who are late talkers will resolve without any extra support.”

Despite numerous studies showing that outcomes of late-talking toddlers are very variable and hard to predict, there are still those who advocate widespread screening of children at age 2 in order to identify potential language delay. This is something that Bishop describes as her “main gripe”.

“Many toddlers who are late talkers do just fine,” she says. “By 4 years, it’s easier to identify those who are likely to have longer-term problems - in fact, prediction at this age is pretty good - and there are some interventions of known effectiveness.”

Measure of success

So, it may be easier and more reliable, at the moment, to pick up problems at school. But one of the major issues with interventions in the under 3 age group and at school age is measurement of effectiveness. Typically, measurement is done by assessing the performance of children of a certain age in areas such as reading or arithmetic, and comparing with the average. But is it really fair to expect children to hit developmental milestones at the same time?

“There is huge variation in normal development,” explains Norbury.

“For example, the age at which children reliably say their first words varies from 355 days to 575 days.

“It does make sense to say, ‘Here are the things that we want all children in Year 3 to be able to do, because these are the foundations for learning in Year 4.’ But we have to accept that there will be variation in when/how children achieve those targets.”

McBride believes that the monitoring of children’s progress is “largely a political choice rather than a scientific one”, although he adds that having a standardised approach to some teaching - such as the introduction of formal reading at age 4 - can have beneficial effects. “What some people would argue is that allowing more flexibility around milestones could create a culture of low expectations,” he says, adding that if teachers were to choose which students should start formal reading at 4, for example, there was a risk that mistakes might be made when deciding who was ready.

“However, it is clearly unrealistic to think that all children, regardless of when they were born in the school year, are going to be at the same point when you do your assessment - that is not going to happen.”

Bishop adds that for some children, it is not easy - or, perhaps, not even possible - to close the language gap completely - so what would our measure of “success” be?

“The best we can do may be to reduce it a bit or even just stop it widening,” she says.

No research consensus

So, while the importance of effective early intervention is not widely disputed, there are numerous examples of targeted language intervention failing to produce the desired results and no real research consensus as to what effective intervention should look like.

If the outcomes are not guaranteed, should we really be dedicating money and time to rolling out widespread programmes of early intervention before we have research-backed interventions that work?

“Many would say, ‘so what? It can’t hurt’, but we don’t often consider the costs,” Norbury says. “For many programmes, parents have to find time out of work...to attend sessions; it might increase their concern about child development and services have limited funds so provision here reduces other types of provision.”

That doesn’t mean that we should never provide intervention to very young children, of course, “but I would target that intervention at young children with multiple risk factors, such as family history, more general developmental concerns [and] significant family concerns”.

She worries there is a danger that if the money is targeted at early age groups, there is a likelihood of budgets being slashed for interventions later.

“I think it is a mistake to believe that, by providing early intervention, we will have no need for intervention or support in the later years,” she says.

And yet, do we stall interventions until we do have the evidence we need or do we start cautiously trying to intervene with the knowledge we already have?

Certainly schools report improvements in their intake by having children in their own pre-school and taking control of the learning that happens there, but that is a judgement call from those involved rather than conclusive evidence.

Clearly, language development is much more complex and far trickier to peg to age-defined milestones than we often seem to believe. And it’s likely that other areas of intervention around developmental milestones will be beset with similar complexity. We can’t expect young children to comply with a neat 0-3 years timeline of development so we can identify and fix problems. And we have a lot of work to do to find what any potential fix might be.

But Bishop stresses that this should not be used as an argument for doing nothing; rather, on the contrary, that we need to be more knowledgeable about early childhood development and more realistic in our expectations.

“I’m all for putting efforts into interventions to boost children’s language skills in the early school years or, indeed, to reduce poverty and make use of things like free school meals,” says Bishop.

“But if history has taught us anything, it’s that there is no quick fix and that people tend to exaggerate the impact of interventions. It’s actually pretty hard to make a big difference, and if you introduce school-based changes, then everyone improves but the [attainment] gap tends to remain.”

Chris Parr is a freelance journalist